© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 18-GEMMASCOPE-RES-16K
    6. 3410
    Prev
    Next
    INDEX
    Explanations

    specific scientific terminology and key terms related to research and analysis in various scientific fields

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    Preceding specific, uncommon words

    np_acts-logits-general · gemini-2.0-flash

    the followed by nouns

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.18.hook_resid_post
    Hook Layer
    18
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     bagus
    -0.71
     buoni
    -0.63
    enumi
    -0.60
     things
    -0.59
     coisas
    -0.56
     Tym
    -0.55
     Oce
    -0.54
    AnchorStyles
    -0.53
     thing
    -0.52
     appunt
    -0.52
    POSITIVE LOGITS
    InitVars
    0.76
    RepeatedField
    0.62
     corresponding
    0.61
    енча
    0.60
     experimentally
    0.59
     externally
    0.58
    ->___
    0.57
    FailureListener
    0.57
     estekak
    0.56
    IndentedString
    0.53
    Activations Density 1.717%

    No Known Activations